Nikon D800, 36 Megapixels and Blur

[jrw]

I just had to try some way to test all of the theorizing that was going on. Much prefer looking at something instead of risking brain strain wrapping my head around concepts and trying to visualize the end results. Understood that resolution of a smaller pixel size would expose lens flaws or resolution capability. Just didn't expect the 100-400 on the 5D to do better than the 70-200 f2.8 IS II on the 7D because I know the 100-400 does not match the resolution of the 70-200 over their shared range. What I didn't check beforehand is that the area of the 7D pixels is around half of the 5D pixels. 70-200 may be better, but not by a factor of 2.

Also looking back at the photos I realized that where I saw off colours in solid colours they were all dark. In lighter areas of the same shots that was no longer present. Dark means less signal reducing SNR values. Design tradeoffs indeed.


@ChadS I must remind you that I said I WON'T be getting rid of my 7D, unless someone will offer me $2000? The resolution for cropping matters to me sometimes as well as use at sports events. Now I'll have a better idea of what the lens and body combos will be able to do. At the end of the day, I have learned something about my tools which I can only hope will improve my usage of them.

[ChadS]

At the end of the day, I have learned something about my tools which I can only hope will improve my usage of them.

Which is why I read this site every day. Glad I may have helped in a small way.

[Kayaker72]

I am just getting around to reading the manual....this line stood out to me:

"With the D800/D800E, you will notice that photos seem to have
less depth of fi eld than pictures shot with other cameras under
the same conditions, and that focus consequently requires more
attention."


This makes sense to me if they crop the image, but I am not sure I follow for an uncropped image. Sorry if this was already covered.

[HDNitehawk]

@Brant; it is almost as if Nikon wrote this guide for people who were upgrading from a crop camera to Full Frame. If you were coming from a crop body you might notice these things. The camera the D800 replaces, the D700, is a full frame camera so in that way it doesn’t make sense. So using a D800 somehow the DOF is going to be narrower than other similar FF bodies?

Of course it says “seem” to have less DOF. If they are trying to say that the high pixel count is going to exaggerate blur then I could see this statement would work. But I just don’t see how this should be, you would think that blur wouldn’t be worse than you would already have out of a D700. It should be at least the same.

[DavidEccleston]

I'm going to pick megapixels numbers that make this easier to visualize. Lets say we had a 1MP camera, and the new camera was 4MP (twice as wide and high). The sensor is the same size (full frame), just more dense.

When light reaches your sensor, everything is a blurred circle, even the things in focus. It's just that the in focus blurred circles are less than a pixel wide. Things that were at the edge of being in focus on the 1MP sensor, ie: the radius was very nearly a full pixel wide, are now 2 pixel wide dots on the 4MP sensor, cleary not in perfect focus.

The "seem" to have less DoF is because it is the same DoF, just you can now more clearly see the limits of the DoF, and because we won't view the images at 1:1, but as say, an 8x10, so all the extra pixels will just smooshed into the same space as always, and we won't notice a difference.

[HDNitehawk]

@David, I think the key word is “seem”. You wouldn’t get an increased area blurred, as the area that was oof is still oof. You would only increase the area that is in focus, which may give you the “seem” effect Nikon mentions. A sharper in focus area gives you a more noticeable distinction between oof areas in focus areas. Would this not be correct under your premise?

We should keep in mind that if Nikon is comparing the D800 to the D700, the D800 has 3x the pixel count.

[neuroanatomist]

I think they mean 1:1 viewing would mean an apparent decrease in DoF. Since we usually normalize (all the online/app calculators do, e.g. an 8x10" print viewed at 30"), there's no objective difference with resolution.

[HDNitehawk]

Since the 5D II is the first full frame digital camera I have owned, I wouldn’t know if this topic has come up before in this way, decreased DOF and more Blur. But it seems to me that when the 5D went from a 12.8mp to a 5D II 21mp it would have been a topic of conversation. Especially if David’s description is accurate in what happens.

[Kayaker72]

Yep...it does seem that they are very paranoid about pixel peepers looking at a 36 MP image. And I am sure it is warranted. I've read comments about noise and different cameras not being sharp, etc and you look into it, someone is looking at an image at 100%...But, if that is what they meant by the statement, I wish they would have make it clear by saying "less depth of field when viewed at 100%"...or something simliar.

[HDNitehawk]

Brant maybe they should have said it like this:

“The D800 is very sharp compared to the old low mp cameras Nikon has made in the past, as a past Nikon user one will now have to practice focus techniques as the new D800 is actually able to produce sharp images with a defined dof. Since one was not able to achieve this level of sharpness in the past it will be an entirely new experience, having images that the bokeh is actually discernable from the in focus areas”

Ok it’s a Nikon flame, I am still waiting on Canon’s response to this new 36mp camera